Web Services – ICWS 2020

Blockchain can achieve non-tampering, non-repudiation, consistency and integrity that other data management technologies do not have. Especially in peer-to-peer networks, the decentralized nature of blockchain has drawn tremendous attention from academic and industrial communities. Recently, the field of e-commerce has also begun to realize its important role. Although blockchain technology has many advantages in achieving trust establishment and data sharing among distributed nodes, in order to make it better to be applied in e-commerce, it is necessary to improve the security of transactions and the efficiency of consensus mechanisms. In this paper, we present a reputation based hybrid consensus to solve the problem of transaction security and efficiency. Our scheme integrates the reputation mechanism into transactions and consensus, and any improper behavior of nodes will be reflected in the reputation system and fed back to a new round of transactions and consensus. We implement distributed reputation management and enable users to append new reputation evaluations to the transaction that has previously evaluated. Meanwhile, we demonstrated that the scheme can defend against existing attacks such as selfish mining attacks, double spending attacks and flash attacks. We implement a prototype and the result shows that our scheme is promising.

As a core technology of the blockchain, the smart contract is receiving increasing attention. However, the frequent outbreak of smart contract security events shows that improving the security of smart contracts is essential. How to guarantee the privacy of contract execution and the correctness of calculation results at the same time is still an issue to be resolved. Using secure multi-party computation (SMPC) technology to implement smart contracts is considered to be one of the potential solutions. But in the existing SMPC based contract execution schemes, a problem has been ignored, that is, the attacker can perform the same process as the reconstructor to recover the secret, which leads to the leakage of users’ privacy. Therefore, in order to solve this problem in the process of smart contract operation, an improved homomorphic encryption algorithm is proposed in this paper, which has a relatively small public key size, short ciphertext length, and high encryption efficiency. Then, a contract execution scheme integrated with SMPC and homomorphic encryption (SMPC-HE for short) is further proposed, which is able to guarantee the privacy of contract execution and the correctness of the calculation results at the same time, and also makes smart contract execution fairer. Finally, our scheme is proved secure, efficient and has low space overhead by theory and experiment results.

The cloud computing paradigm is characterized by the ability to provide flexible provisioning patterns for computing resources and on-demand common services. As a result, building business processes and workflow-based applications on cloud computing platforms is becoming increasingly popular. However, since real-world cloud services are often affected by real-time performance changes or fluctuations, it is difficult to guarantee the cost-effectiveness and quality-of-service (Qos) of cloud-based workflows at real time. In this work, we consider that workflows, in terms of Directed Acyclic Graphs (DAGs), to be supported by decentralized cloud infrastructures are with time-varying performance and aim at reducing the monetary cost of workflows with the completion-time constraint to be satisfied. We tackle the performance-fluctuation workflow scheduling problem by incorporating a stochastic-performance-distribution-based framework for estimation and optimization of workflow critical paths. The proposed method dynamically generates the workflow scheduling plan according to the accumulated stochastic distributions of tasks. In order to prove the effectiveness of our proposed method, we conducted a large number of experimental case studies on real third-party commercial clouds and showed that our method was significantly better than the existing method.

Recently, the developer recommendation on crowdsourcing software platform is of great research significance since an increasingly large number of tasks and developers have gathered on the platforms. In order to solve the problem of cold-start, the existing developer recommendation algorithms usually only use explicit information but not ID information to represent tasks and developers, which causes poor performance. In view of the shortcomings of the existing developer recommendation algorithms, this paper proposes an FM recommendation algorithm based on explicit to implicit feature mapping relationship modeling. This algorithm firstly integrates fully the ID information, explicit information and rating interaction between the completed task and the existing developers by using FM algorithm in order to get the implicit features related to their ID information. Secondly, for the completed tasks and existing developers, a deep regression model is established to learn the mapping relationship from explicit features to implicit features. Then, for the cold-start task or the cold-start developer, the implicit features are determined by the explicit features according to the deep regression model. Finally, the ratings in the cold-start scene can be predicted by the trained FM model with the explicit and implicit features. The simulation results on Topcoder platform show that the proposed algorithm has obvious advantages over the comparison algorithm in precision and recall.

As the emerging entertainment applications, short video platforms, such as Youtube, Kuaishou, quickly dominant the Internet multimedia traffic. The caching problem will surely provide a great reference to network management (e.g., traffic engineering, content delivery). The key to cache is to make precise popularity prediction. However, different from traditional multimedia applications, short video network exposes unique characteristics on popularity prediction due to the explosive video quantity and the mutual impact among these countless videos, making the state-of-the-art solutions invalid. In this paper, we first give an in-depth analysis on 105,231,883 real traces of 12,089,887 videos from Kuaishou Company, to disclose the characteristics of short video network. We then propose a graph convolutional neural-based video popularity prediction algorithm called GraphInf. In particular, GraphInf clusters the countless short videos by region and formulates the problem in a graph-based way, thus addressing the explosive quantity problem. GraphInf further models the influence among these regions with a customized graph convolutional neural (GCN) network, to capture video impact. Experimental results show that GraphInf outperforms the traditional Graph-based methods by 44.7%. We believe such GCN-based popularity prediction would give a strong reference to related areas.

Crowdsourcing is a way to solve problems that need human contribution. Crowdsourcing platforms distribute replicated tasks to workers, pay them for their contribution, and aggregate answers to produce a reliable conclusion. A fundamental problem is to infer a correct answer from the set of returned results. Another challenge is to obtain a reliable answer at a reasonable cost: unlimited budget allows hiring experts or large pools of workers for each task but a limited budget forces to use resources at best.

This paper considers crowdsourcing of simple boolean tasks. We first define a probabilistic inference technique, that considers difficulty of tasks and expertise of workers when aggregating answers. We then propose CrowdInc, a greedy algorithm that reduces the cost needed to reach a consensual answer. CrowdInc distributes resources dynamically to tasks according to their difficulty. We show on several benchmarks that CrowdInc achieves good accuracy, reduces costs, and we compare its performance to existing solutions.

In recent years, Web Application Programming Interfaces (APIs) are becoming more and more popular with the development of the Internet industry and software engineering. Many companies provide public Web APIs for their services, and developers can greatly accelerate the development of new applications by relying on such APIs to execute complex tasks without implementing the corresponding functionalities themselves. The proliferation of web APIs, however, also introduces a challenge for developers to search and discover the desired API and its endpoint. This is a practical and crucial problem because according to ProgrammableWeb, there are more than 22,000 public Web APIs each of which may have tens or hundreds of endpoints. Therefore, it is difficult and time-consuming for developers to find the desired API and its endpoint to satisfy their development needs. In this paper, we present an intelligent system for Web API searches based on natural language queries by using a two-step transfer learning. To train the model, we collect a significant amount of sentences from crowdsourcing and utilize an ensemble deep learning model to predict the correct description sentences for an API and its endpoint. A training dataset is built by synthesizing the correct description sentences and then is used to train the two-step transfer learning model for Web API search. Extensive evaluation results show that the proposed methods and system can achieve high accuracy to search a Web API and its endpoint.

Anti-patterns are symptoms of poor design and implementation solutions applied by developers during the development of their software systems. Recent studies have identified a variety of Web service anti-patterns and defined them as sub-optimal solutions that result from bad design choices, time pressure, or lack of developers experience. The existence of anti-patterns often leads to software systems that are hard to understand, reuse, and discover in practice. Indeed, it has been shown that service designers and developers tend to pay little attention to their service interfaces design. Web service antipatterns detection is a non-trivial and error-prone task as different anti-pattern types typically have interleaving symptoms that can be subjectively interpreted and hence detected in different ways. In this paper, we introduce an automated approach that learns from a set of interleaving Web service design symptoms that characterize the existence of anti-pattern instances in a service-based system. We build a multi-label learning model to detect 8 common types of Web service anti-patterns. We use the ensemble classifier chain (ECC) model that transforms multi-label problems into several single-label problems which are solved using genetic programming (GP) to find the optimal detection rules for each anti-pattern type. To evaluate the performance of our approach, we conducted an empirical study on a benchmark of 815 Web services. The statistical tests of our results show that our approach can detect the eight Web service antipattern types with an average F-measure of 93% achieving a better performance compared to different state-of-the-art techniques. Furthermore, we found that the most influential factors that best characterize Web service anti-patterns include the number of declared operations, the number of port types, and the number of simple and complex types in service interfaces.

E-Government services are persistent targets of the organized crime by hackers, which hinders the delivery of services. Computational trust is an important technique for the security work of service providers (SPs). However, it relies on data collection about users’ past behaviors conventionally from other SPs, which incurs the uncertainty of data and thus impacts the quality of data. Motivated by this issue, this paper proposes a novel smart contract based user-centric computational trust framework (UCCT) which collects the behavioral data of the user. It uses smart contract as a rational trustworthy agent to automatically monitor and manage the user’s behaviors on the user side, so as to provide deterministic data quality assurance services for the computational trust. Furthermore, a privacy-preserving way of the data sharing is provided for the user and a personalized security mechanism for the SP. A new ledger is also introduced to provide a user-centric and efficient search. The results of experiments conducted on a Hyperledger Fabric based blockchain platform demonstrate that the time cost of user-centric ledger in UCCT can be less than 1 s. Moreover, even if a more complicated contract is provided, the improvement of transaction per second (TPS), which is made by UCCT, is not less than 8%.

Twitter is a popular social networking platform that is widely used in discussing and spreading information on global events. Twitter trending hashtags have been one of the topics for researcher to study and analyze. Understanding the posting behavior patterns as the information flows increase by rapid events can help in predicting future events or detection manipulation. In this paper, we investigate similar-context trending hashtags to characterize general behavior of specific-trend and generic-trend within same context. We demonstrate an analysis to study and compare such trends based on spatial, temporal, content, and user activity. We found that the characteristics of similar-context trends can be used to predict future generic trends with analogous spatiotemporal, content, and user features. Our results show that more than 70% users participate in location-based hashtag belongs to the location of the hashtag. Generic trends aim to have more influence in users to participate than specific trends with geographical context. The retweet ratio in specific trends is higher than generic trends with more than 79%.

Scholarly digital libraries provide access to scientific publications and comprise useful resources for researchers who search for literature on specific subject areas. CiteSeerX is an example of such a digital library search engine that provides access to more than 10 million academic documents and has nearly one million users and three million hits per day. Artificial Intelligence (AI) technologies are used in many components of CiteSeerX including Web crawling, document ingestion, and metadata extraction. CiteSeerX also uses an unsupervised algorithm called noun phrase chunking (NP-Chunking) to extract keyphrases out of documents. However, often NP-Chunking extracts many unimportant noun phrases. In this paper, we investigate and contrast three supervised keyphrase extraction models to explore their deployment in CiteSeerX for extracting high quality keyphrases. To perform user evaluations on the keyphrases predicted by different models, we integrate a voting interface into CiteSeerX. We show the development and deployment of the keyphrase extraction models and the maintenance requirements.

The edge computing paradigm is featured by the ability to off-load computing tasks from mobile devices to edge clouds and provide high cost-efficient computing resources, storage and network services closer to the edge. A key question for workflow scheduling in the edge computing environment is how to guarantee user-perceived quality of services when the supporting edge services and resources are with unstable, time-variant, and fluctuant performance. In this work, we study the workflow scheduling problem in the multi-user edge computing environment and propose a Deep-Q-Network (DQN) -based multi-workflow scheduling approach which is capable of handling time-varying performance of edge services. To validate our proposed approach, we conduct a simulative case study and compare ours with other existing methods. Results clearly demonstrate that our proposed method beats its peers in terms of convergence speed and workflow completion time.

Content caching in mobile edge networks has stirred up tremendous research attention. However, most existing studies focus on predicting content popularity in mobile edge servers (MESs). In addition, they overlook how the content is cached, especially how to cache the content with user devices. In this paper, we propose CMU, a three-layer (Cloud-MES-Users) content caching framework and investigate the performance of different caching strategies under this framework. A user device who has cached the content can offer the content sharing service to other user devices through device-to-device communication. In addition, we prove that optimizing the transmission performance of CMU is an NP-hard problem. We provide a solution to solve this problem and describe how to calculate the number of distributed caching nodes under different parameters, including time, energy and storage. Finally, we evaluate CMU through a numerical analysis. Experiment results show that content caching with user devices could reduce the requests to Cloud and MESs, and decrease the content delivery time as well.